Search

Background Routine checking of gastric residual volumes (GRVs) during enteral feeding within surgical trauma intensive care units (STICUs) is a common practice. However, data on the necessity of this practice and its impact on nutrient delivery are limited. We aim to study the association between the replacement of a routine GRV (rGRV) policy with a triggered GRV (tGRV) policy and the safe achievement of daily nutrition goals. Methods We prospectively collected data on patients after we instituted a tGRV policy and compared them with a historical cohort of patients who had rGRV assessments in our STICU at a level 1 trauma center. The primary end point was achieving 80% of prescribed nutrient goals. Secondary end points included aspiration pneumonia, witnessed emesis, and glycemic control. Results A total of 145 patients accounting for 1405 STICU days were treated under the tGRV policy, and 156 patients accounting for 1694 STICU days were treated under the rGRV policy. There were no statistically significant differences between the tGRV and rGRV groups with regard to the proportion of days meeting or exceeding protein (56.7% vs 56.2%) or calorie (56.4% vs 56.0%) goals. After adjusting for in-hospital deaths, injury severity score, complications, and STICU time, the predictive margins for meeting caloric and protein goals were higher among the tGRV patients (57% vs 56%), but these differences were not statistically significant. Conclusion A tGRV policy did not change protein or calorie delivery among patients or increase the risk of emesis compared with traditional monitoring methods.

Background: Cervical spine fractures are common in traumatically injured patients. The halo-vest brace is a common treatment used for these fractures. We hypothesize that the use of halo-vest fixation is associated with a high incidence of dysphagia in trauma patients. Methods: All trauma patients at our Level I Trauma Center from August 2005 to August 2007 were analyzed retrospectively via the trauma registry (N = 3,702). Included were adult patients with cervical spine fractures treated with halo-vests and evaluated formally by speech-language pathologists for dysphagia and aspiration. Patients were categorized into mild, moderate, and severe dysphagia. Results: Of the 3,702 patients, 369 (10%) had cervical spine fractures from blunt trauma and 56 met inclusion criteria. Of these, 19 (34%) had no evidence of swallowing dysfunction and the remaining 37 (66%) had evidence of dysphagia. Thirteen (23%) exhibited symptoms of aspiration. There were no significant differences in age, gender, Injury Severity Score, arrival Revised Trauma Score, or arrival Glasgow Coma Scale score on presentation. Dysphagia is associated with longer intensive care unit stays (p = 0.019) and trends toward a longer hospital stay (p = 0.083). In trauma patients with halo-vests, increasing severity of dysphagia from mild to moderate is associated with longer ventilator days (p = 0.005), intensive care unit days (p = 0.001), and hospital length of stay (p = 0.015). Conclusions: Patients with cervical fractures treated with halo-vest fixation have a significantly high incidence of dysphagia and aspiration. Dysphagia in trauma patients treated with halo-vests for c-spine fractures is common, associated with worse outcomes, and difficult to predict. Therefore, all of these patients should be formally evaluated for dysphagia.

Skin flora immobilization technology is similar in efficacy to Iodine-Povidone in healthy volunteers. We did a prospective study in a university clinic with 60 healthy volunteers. Right inguinal skin area on healthy volunteers was used to compare the antimicrobial properties of cyanoacrylate sealant (FloraSeal, Adhesion Biomedical, Wyomissing, PA) versus standard surgical preparation Povidone-iodine (Betadine, Purdue Productions, Stamford, CT). Bacterial counts were measured at different time intervals: 15 minutes, 4 hours, and 24 hours. Bacterial colony forming units were compared between Povidone-iodine and cyanoacrylate sealant. The absolute log reduction was 5.568 for Povidone-iodine (7 absolute CFU); 5.028 for cyanoacrylate (59 absolute CFU); and 5.568 for Povidone-iodine and cyanoacrylate combined (21 absolute CFU). Cyanoacrylate was able to sustain a reduction on bacterial counts at 4 hours and 24 hours of more than 99.8 per cent as compared with the control group. Cyanoacrylate microbial sealant successfully reduces bacterial counts on normal healthy skin. The results were similar to Povidone-iodine alone. We believe this technology may be an excellent means of mitigating incisional surgical site infection by reducing the risk of contamination by skin flora and warrants further testing.

We performed a retrospective chart review of trauma patients admitted to Palmetto Richland Memorial Hospital and identified 63 cases of adrenal insufficiency along with 65 trauma patient controls. Two statistical models, a neural network and a multiple logistic regression, were developed to predict patients with increased risk of developing adrenal insufficiency. Each model had 11 selected independent variables, along with patient demographic data, to make a probabilistic prediction of patient outcome. The neural network model was trained with 102 patients to identify linear and nonlinear relationships that could yield a predictive capability. The neural network achieved an accuracy of 71 per cent. The logistic regression model achieved an accuracy of 82 per cent. With these models, we have shown the feasibility of a method to more accurately screen patients with an increased risk of adrenal insufficiency. This ability should allow earlier identification and treatment of patients with adrenal insufficiency. Further development with a larger database is needed to improve the accuracy of the present models.

Diabetes insipidus (DI) is an uncommon but important complication in the head-injured population. A retrospective review of all trauma patients admitted to the intensive care unit (ICU) during a 4-year period who developed DI was undertaken. The incidence of DI was 1.3 per cent in ICU trauma admissions and 2.9 per cent in traumatic head injuries admitted to the ICU. The overall mortality was 69 per cent (18/26). The mean onset time of DI in nonsurvivors (1.5 ± 0.7 days) was shorter compared to survivors (8.9 ± 10.2 days) ( P < 0.001). All patients who died developed DI within the first 3 days of hospitalization. Patients who develop DI early in their course have a higher mortality than those who develop DI later in their hospital course. The development of diabetes insipidus after head injury carries a 69 per cent mortality rate, and if the onset is within the first 3 days after injury, mortality rate rises to 86 per cent.

The inability for abdominal closure in critically ill surgical patients provides a complex problem. Often, these patients are left with a large ventral hernia, which requires readmission for abdominal wall repair. We are reporting on the use of a vacuum-assisted device (VAD) to facilitate abdominal wall closure. Fifteen patients were enrolled for placement of a VAD. Selection was based on the diagnosis of abdominal compartment syndrome, the inability for abdominal closure at the initial operation, or the inability to close the abdomen upon re-exploration. Ten (67%) patients were successfully closed within 11 days using the VAD. Predictors of successful closure were the duration of VAD placement (

Purpose: The surgical resident experience with trauma has changed. Many residents are exposed to predominantly nonoperative patient care experiences while on trauma rotations. Data from a large multicenter study were analyzed to estimate surgical resident exposure to trauma laparotomy, diagnostic peritoneal lavage (DPL), and focused abdominal sonography for trauma (U/S). Methods: Centers completed a self-report questionnaire on their institutional demographics, admissions, and procedure for a 2-year period (1998-1999). Results: A total of 82 trauma centers that provide resident teaching were included. The included centers represent over 247,000 trauma admissions. The majority of trauma centers (65.9%) had > 80% blunt injury. Although all centers performed laparotomies, other results were more variable. For U/S, 24.2% performed none at all and 47.0% performed fewer than two U/S examinations per month. For DPLs, 3.8% performed none and 66.7% performed fewer than two per month. Assuming 1 night of 4 on call, the average surgical resident training at a trauma center performing > 80% blunt trauma has the potential to participate in only 15 trauma laparotomies, 6 diagnostic peritoneal lavages, and 45 ultrasound examinations per year. In addition, the resident will care for an average of 500 blunt trauma patients before performing a splenectomy or liver repair. Conclusion: Surgical resident experience on most trauma services is heavily weighted to nonoperative management, with a relatively low number of procedures, little experience with DPL, and highly variable experience with ultrasound. These data have serious implications for resident training and recruitment into the specialty.

Incisional hernias after abdominal operations are a significant cause of long-term morbidity and have been reported to occur in 3 to 20 per cent of laparotomy incisions. Traditional primary suture closure repair is plagued with up to a 50 per cent recurrence rate. With the introduction of prosthetic mesh repair recurrence decreased, but complications with mesh placement emerged ushering in the development of laparoscopic incisional herniorrhaphy. The records of patients who underwent laparoscopic incisional hernia repair between June 1, 1995 and September 1, 2001 were reviewed. Patient demographics, hernia defect size, recurrence, operative time, and procedure-related complications were evaluated. Fifty patients (22 male and 28 female, mean age 57 years with range of 24–83) were scheduled for laparoscopic incisional hernia repair between June 1, 1995 and September 1, 2001. The average patient was obese with a mean body mass index of 35.8 kg/m2(range 16–57 kg/m2). Two patients (4%) had primary ventral hernias. Forty-eight patients (96%) had incisional hernias with 22 (46%) of these previously repaired with prosthetic mesh. Mean defect size was 206.1 cm2(range 48–594 cm2). The average mesh size was 510.2 cm2(range 224–1050 cm2). Gore-Tex® DualMesh and Bard® Composite Mesh were used in 84 and 16 per cent of the repairs, respectively. Mean operating time was 97 minutes. There were no deaths. Complications were seen in 12 per cent patients (six occurrences) and included two small bowel enterotomies, a symptomatic seroma requiring aspirate, a mesh reaction requiring a short course of intravenous antibiotics, and trocar site pain (two patients). There were no recurrences during a mean follow-up of 41 months (range 3–74 months). We conclude that laparoscopic incisional herniorrhaphy offers a safe and effective repair for large primary and recurrent ventral hernia with low morbidity.

The efficacy trial of diaspirin cross-linked hemoglobin (DCLHb) in traumatic hemorrhagic shock demonstrated an unexpected mortality imbalance, prompting a three-step review to better understand the cause of this finding.Patients were enrolled in this DCLHb hemorrhagic shock study using 28-day mortality as the primary endpoint. Mortality data were primarily analyzed using the TRISS method and a nonblinded clinical review, followed by an independent Pennsylvania Trauma Outcome Study (PTOS)-derived probability of survival analyses. Finally, a trauma expert conducted a blinded clinical review of cases incorrectly predicted by these PTOS analyses.More of the DCLHb patients predicted to survive using TRISS actually died than in the control subgroup (24% vs. 3%, p0.002). Nonblinded clinical review noted that 72% of the patients who died had prior traumatic arrest, a presenting Glasgow Coma Scale score of 3, or a base deficit15 mEq/L. DCLHb patients predicted to survive using PTOS also more often died than did control patients (30% vs. 8%, p0.04). Blinded clinical review determined that 94% of the deaths were clinically justified. Both the TRISS and the PTOS models gave an adjusted mortality relative risk of 2.3, similar to the unadjusted risk data.Mortality analysis in this shock study involved both clinical case reviews and mortality prediction models. Despite the observation that nearly all of the deaths were clinically justified, the TRISS and PTOS models demonstrated excess unpredicted deaths in the DCLHb subgroup. A combined process, using both mortality prediction models and clinical case reviews, is useful in trauma studies that use a mortality endpoint.

Background: Current literature suggests that blunt carotid injuries (BCIs) and vertebral artery injuries (BVIs) are more common than once appreciated. Screening criteria have been suggested, but only one previous study has attempted to identify factors that predict the presence of BCI/BVI. This current study was conducted for two reasons. First, we wanted to determine the incidence of BCI/ BVI in our institution. Second, we wanted to determine the incidence of abnormal four-vessel cerebral angiograms ordered for injuries and signs believed to be associated with BCI/BVI and thus to determine whether the screening protocol developed was appropriate. Methods: From August 1998, we used liberalized screening criteria for patients who were prospectively identified and suspected to be at high risk for BCI/ BVI if any of the following were present: anisocoria, unexplained mono-/hemiparesis, unexplained neurologic exam, basilar skull fracture through or near the carotid canal, fracture through the foramen transversarium, cerebrovascular accident or transient ischemic attack, massive epistaxis, severe flexion or extension cervical spine fracture, massive facial fractures, or neck hematoma. Four-vessel cerebral angiograms were used for screening for BCI/BVI. Results: Over the 18-month study period, 48 patients were angiographically screened, with 21 patients (44%) being identified as having a total of 19 BCIs and 10 BVIs. Nine patients had unilateral carotid artery injuries and three patients had bilateral carotid artery injuries. Vertebral artery injuries were unilateral in six patients. One patient had bilateral carotid artery injuries and a unilateral vertebral artery injury. One patient had a unilateral carotid artery injury and a unilateral vertebral artery injury, and one patient had a unilateral carotid artery injury and bilateral vertebral artery injuries. During the same study period, 2,331 trauma patients were admitted, with 1,941 (83%) secondary to blunt trauma. The overall incidence of BCI/BVI was 1.1%. The frequency of abnormal angiograms ordered for cerebrovascular accident or transient ischemic attack, massive epistaxis, or severe cervical spine fractures was 100%. The frequency of abnormal angiograms ordered for the other indications was as follows: fracture through foramen transversarium, 60%; unexplained mono- or hemiparesis, 44%; basilar skull fracture, 42%; unexplained neurologic examination, 38%; anisocoria, 33%; and severe facial fractures, 0%. Conclusion: The liberalized screening criteria used in this study were appropriate to identify patients with BCI/BVI. This study suggests BCI/BVI to be more common than previously believed and justifies that screening should be liberalized.

(2000). Hemoglobin Solutions as Blood Substitutes. Journal of Investigative Surgery: Vol. 13, No. 2, pp. 79-94.

Skin flora immobilization technology is similar in efficacy to Iodine-Povidone in healthy volunteers. We did a prospective study in a university clinic with 60 healthy volunteers. Right inguinal skin area on healthy volunteers was used to compare the antimicrobial properties of cyanoacrylate sealant (FloraSeal, Adhesion Biomedical, Wyomissing, PA) versus standard surgical preparation Povidone-iodine (Betadine, Purdue Productions, Stamford, CT). Bacterial counts were measured at different time intervals: 15 minutes, 4 hours, and 24 hours. Bacterial colony forming units were compared between Povidone-iodine and cyanoacrylate sealant. The absolute log reduction was 5.568 for Povidone-iodine (7 absolute CFU); 5.028 for cyanoacrylate (59 absolute CFU); and 5.568 for Povidone-iodine and cyanoacrylate combined (21 absolute CFU). Cyanoacrylate was able to sustain a reduction on bacterial counts at 4 hours and 24 hours of more than 99.8 per cent as compared with the control group. Cyanoacrylate microbial sealant successfully reduces bacterial counts on normal healthy skin. The results were similar to Povidone-iodine alone. We believe this technology may be an excellent means of mitigating incisional surgical site infection by reducing the risk of contamination by skin flora and warrants further testing.

Duplex ultrasonography was used prospectively in the initial evaluation of 198 patients with 319 potential vascular injuries of the neck and extremities. Patients who were unstable or who had obvious arterial trauma were excluded. Injury was caused by gunshot in 104 (53%), blunt trauma in 42 (21%), stab wound in 34 (17%), and shotgun in 18 (9%). Duplex ultrasonography correctly characterized and localized vascular injuries in 23 patients: arterial disruptions (13), intimal flaps (4), acute pseudoaneurysms (3), arteriovenous fistulas (2), and shotgun pellet arteriopuncture (1). Nineteen other patients had vasospasm (13) or external compression (6) without evidence of intrinsic vessel injury, these 42 studies had true-positive results. Twenty patients underwent arterial repair (13 on the basis of duplex ultrasonography alone), one had primary amputation, three required fasciotomy, and 18 were observed. Two patients with false-negative results had minor shotgun pellet arteriopunctures that were missed by duplex ultrasonography, but neither needed repair. One hundred fifty-three patients had true-negative results on duplex ultrasonography: all clinically had only proximity injuries and easily palpable distal pulses. The result of one duplex ultrasonography study was found to be false-positive on arteriography. The sensitivity of duplex ultrasonography was 95%, the specificity was 99%, and the overall accuracy was 98%. These results closely approximate those reported with the use of exclusion arteriography in the evaluation of similar vascular trauma patients. Furthermore, duplex ultrasonography has no interventional risks and is more cost-effective for screening such injuries than arteriography or exploration. Duplex ultrasonography is a reliable method of diagnosis in patients with potential peripheral vascular injuries.

The inability for abdominal closure in critically ill surgical patients provides a complex problem. Often, these patients are left with a large ventral hernia, which requires readmission for abdominal wall repair. We are reporting on the use of a vacuum-assisted device (VAD) to facilitate abdominal wall closure. Fifteen patients were enrolled for placement of a VAD. Selection was based on the diagnosis of abdominal compartment syndrome, the inability for abdominal closure at the initial operation, or the inability to close the abdomen upon re-exploration. Ten (67%) patients were successfully closed within 11 days using the VAD. Predictors of successful closure were the duration of VAD placement (12 days, P0.001), the total amount of VAD output (3 L, P0.04), the patient's cumulative fluid balance within the first 2 weeks (2 L, P0.002), or the presence of a systemic infection at the time of attempted closure (P0.001). After 6 months, there have been no complications in patients successfully closed with this device. There have been a few recent reports describing VAD abdominal closures. While not successful for every case, the majority of our patients were able to have their abdominal wall closed primarily. We plan to use this technique to help shorten hospital stay and prevent readmission for hernia repair.

There are many reasons for hypotension in trauma patients with multiple injuries; one uncommon source is facial fractures. The treatment algorithm is volume replacement and local control of the bleeding. A retrospective study was undertaken to evaluate the treatment of patients with life-threatening hemorrhage secondary to facial fractures, and to develop a treatment algorithm.A retrospective chart review was undertaken to determine the incidence of hemorrhagic shock in patients with facial fractures exclusive of others sources, and the use of transcatheter arterial embolization to control the bleeding was evaluated.Over a 4-year period, 7562 patients were treated at Palmetto Richland Memorial Hospital, a Level I trauma center. There were 912 patients with facial injuries, with 11 of these patients presenting with life-threatening hemorrhage secondary to facial fractures. The incidence of life-threatening hemorrhage from facial fracture was 1.2%. The mechanism of injury was blunt in 10 patients and penetrating in 1. The blunt injuries resulted from six motor vehicles crashes, three motorcycle crashes, and one plane crash. The one penetrating injury was a shotgun blast. There were six patients with Le Fort III fractures, two patients with Le Fort II fractures, and three patients with a combination of Le Fort II and III fractures bilaterally. The average volume infused before the embolization was 7 L; this included blood and crystalloid. There were four complications: two minor groin hematomas, one partial necrosis of the tongue, and one facial nerve palsy. There were two deaths, both secondary to concomitant intracranial injury as a result of blunt trauma.The incidence of severe hemorrhage secondary to facial fractures is rare; however, it can be life threatening. When common modalities of treatment such as pressure, packing, and correction of coagulopathy fail to control the hemorrhage, transcatheter arterial embolization offers a safe alternative to surgical control.

Hypothesis: Factors associated with fetal death in injured pregnant patients are related to increasing injury severity and abnormal maternal physiologic profile. Design: A multi-institutional retrospective study of 13 level I and level II trauma centers from 1992 to 1996. Main Outcome Measure: Fetal survival. Results: Of 27 715 female admissions, there were 372 injured pregnant patients (1.3%); 84% had blunt injuries and 16% had penetrating injuries. There were 14 maternal deaths (3.8%) and 35 fetal deaths (9.4%). The population suffering fetal death had higher injury severity scores (P,.001), lower Glascow Coma Scale scores (P,.001), and lower admitting maternal pH (P = .002). Most women who lost their fetus arrived in shock (P = .005) or had a fetal heart rate of less than 110 beats/min at some time during their hospitalization (P,.001). An Injury Severity Score greater than 25 was associated with a 50% incidence of fetal death. Placental abruption was the most frequent complication, occurring in 3.5% of patients and associated with 54% mortality. Cardiotrophic monitoring to detect potentially threatening fetal heart rates was performed on only 61% of pregnant women in their third trimester. Of these patients, 7 had abnormalities on cardiotrophic monitoring and underwent successful cesarean delivery. Conclusions: Fetal death was more likely with greater severity of injury. Cardiotrophic monitoring is underused in injured pregnant patients in their third trimester even after admission to major trauma centers. Increased use of cardiotrophic monitoring may decrease the mortality caused by placental abruption.

Background: Blunt aortic injury is a major cause of death from blunt trauma. Evolution of diagnostic techniques and methods of operative repair have altered the management and posed new questions in recent years. Methods: This study was a prospectively conducted multicenter trial involving 50 trauma centers in North America under the direction of the Multi-institutional Trial Committee of the American Association for the Surgery of Trauma. Results: There were 274 blunt aortic injury cases studied over 2.5 years, of which 81% were caused by automobile crashes. Chest computed tomography and transesophageal echocardiography were applied in 88 and 30 cases, respectively, and were 75 and 80% diagnostic, respectively. Two hundred seven stable patients underwent planned thoracotomy and repair. Clamp and sew technique was used in 73 (35%) and bypass techniques in 134 (65%). Overall mortality was 31%, with 63% of deaths being attributable to aortic rupture; mortality was not affected by method of repair. Paraplegia occurred postoperatively in 8.7%. Logistic regression analysis demonstrated clamp and sew (p = 0.002) and aortic cross clamp time of 30 minutes (p = 0.01) to be associated with development of postoperative paraplegia. Conclusions: Rupture after hospital admission remains a major problem. Although newer diagnostic techniques are being applied, at this time aortography remains the diagnostic standard. Aortic cross clamp time beyond 30 minutes was associated with paraplegia; bypass techniques, which provide distal aortic perfusion, produced significantly lower paraplegia rates than the clamp and sew approach.

Patients with penetrating pericardial trauma whose vital signs stabilize after fluid administration may present a therapeutic dilemma. Two-dimensional echocardiography has emerged as a diagnostic technique to help determine whether surgical intervention may be required. We present 5 patients with penetrating pericardial trauma whose vital signs stabilized after fluid administration and who had minimal clinical findings. In 3 of these patients, a small effusion was seen; in 2 others. no abnormalities were noted. All 5 underwent surgical exploration and had major intrapericardial injuries. We conclude that a normal echocardiographic study does not rule out major intrapericardial injury in patients with penetrating chest trauma. Furthermore, small areas of effusion seen on echocardiography in these patients represent indications for surgical exploration.

Hepatic failure is frequently seen following severe hemorrhagic shock, sepsis, and trauma. Clearance of various drugs has been used to evaluate hepatocellular dysfunction, including indocyanine green (ICG), an organic anionic dye that is transported similarly to bilirubin, and antipyrine (AP), a marker of oxidative phosphorylation. Previous investigators have noted a decrease in ICG excretion following systemic hemorrhage. The effect of isolated hepatic ischemia on the clearances of ICG and AP was studied in 16 pigs after 90 minutes of vascular occlusion to the liver. Antipyrine clearance decreased almost 50% from baseline values at 24 and 72 hours after the ischemia procedure, indicating a significant depression in the cytochrome P-450 system. On the other hand, ICG clearance did not change significantly. In conclusion, ICG clearance is not depressed after isolated hepatic ischemia in pigs. Changes in organic anion clearance after systemic hemorrhage may be because of release of toxic products from ischemic peripheral tissue.

Few, if any, complications have been reported with the nonoperative management of selected hepatic injuries diagnosed by computed tomographic (CT) scan in hemodynamically stable patients. This retrospective study was designed to evaluate complications associated with this form of management. Twenty-six patients (21%) of 128 patients with blunt hepatic injuries were treated nonoperatively over a 3-year period. All patients were hemodynamically stable at the time of admission and had hepatic injuries identified by CT scans of the abdomen. Five patients (19%) developed complications associated with nonoperative therapy. Of these, two patients had minor hepatic injuries (grades 1-2) and three had major (grades 3-5) hepatic injuries. Two patients (one with minor and one with major hepatic injury), developed free intraperitoneal biliary leaks and required operative repair. Three patients (one with minor and two with major hepatic injuries) developed large subcapsular bilomas with resultant hepatic dysfunction. These patients were successfully managed with percutaneous CT-guided drainage. There were no deaths in our study population with nonoperative therapy. The complications of hepatic injuries initially managed by expectant observation were treated operatively or by percutaneous CT-guided drainage. Repeated CT evaluation to follow the progress of liver fracture and the occasional use of hepatobiliary scans for the identification of biliary leaks have proven useful in our experience.

Septic complications following traumatic injury continue to be a contributing factor to morbidity and mortality. Paranasal sinusitis is being recognized as an often occult etiology of fever and sepsis in multiply injured patients. Our series of 11 patients who developed clinically important maxillary sinusitis is presented. Common features of the patients include: 1) nasal instrumentation; 2) craniofacial trauma; 3) concomitant use of steroids; and 4) severe multisystem injury (mean I.S.S., 45.5; T.S., 10.6). A high index of suspicion in patients with nasal tubes who develop unexplained fever or signs of systemic sepsis should prompt appropriate investigation of the paranasal sinuses. Removal of the tubes, antral puncture for irrigation and aspiration for microbiologic culture, topical nasal decongestants, systemic antibiotics based on sensitivity studies, and occasionally, formal surgical sinus drainage contribute to effective therapy.